LXXXIV. T. W. KEELE. 
seriously fallen off, and must therefore conclude that the 
water table must be low and is still falling. 
The Melbourne rainfall diagram (Plate 3) shows that there 
has been an accumulated loss of rain below the mean at 
that place since the year 1875, amounting to 47°94 inches, 
or 1°8 years mean rainfall, and although a much higher 
rainfall is experienced on the catchment areas (probably 
nearly double that of Melbourne) it is reasonable to suppose 
that the same influences which have resulted in the decline 
at Melbourne, have affected those areas in a similar manner 
and the water table there is in all probability at a lower 
level than ever previously known. 
In view, however, of the large quantity of water still 
running from a comparatively small area like the Watts’ 
River catchment, viz., 553 square miles, it is evident that 
there must be an enormous underground storage, either in 
the upper or lower portions of the area, probably in the 
latter. It would thus appear to be feasible, in order to 
maintain the supply in the aqueduct, or to augment it if 
necessary, to drain the water out from the saturated beds by 
making cuttings or trenches into the hill sides, or by con- 
structing drains to be laid in trenches and parallel to the 
creek beds, properly protected with broken stone surround- 
ing the pipes in order to arrest the sand while permitting 
the water to enter the joints or perforations. In thismanner 
no doubt as much water as may be required could be liber- 
ated from the saturated ground by extending the drains. 
While this system would no doubt be very effective in keep- 
ing up the supply, it would assist in lowering the level of 
the water table in those areas where the operations were 
carried on, and it would therefore be necessary to construct 
works in suitable places in the upper portions of the catch- 
ment area to impound water resulting from heavy rains, — 
which would exceed the discharging capacity of theaqueduct — 
and would consequently run to waste. 
